Elecraft

Need help for sick K1

Classic

List

Threaded

31 messages Options

Rick Dettinger-2

Need help for sick K1

When I built my (second) K1-4 last year, it worked fine in all respects, with output as expected. Couple months ago it turned into a K1-2. I got full power on 40 and 30 M but less than 1/2 W on 20 and 15 M. I took it apart and measured RF voltages on 20 M. They were low at all points. I put it together without the tuner. It worked on all bands. A K1-4 again. I put tuner back and it still worked fine. I used it without problem until today. Bact to a K1-2 again. 20 and 15 M had less than 1/2 W. I took it apart and removed tuner again. Big change. Now I get too much power on all bands. 10 W on 40 M to 5 W on 15 M. I have never got more than 7 W on any band. I turned power down to 0.1 W and still got high power on all bands. Also, the RF indicator on the K1 dosn't work. No bars light. Looks like I have a power control problem. Like no control. Any ideas? Supply voltage was 12.7 V.
73
Rick
K7MW
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Don Wilhelm-3

RE: Need help for sick K1

Rick,

Look carefully, then look again and again for the bad solder connection in
the K1 tuner.

73,
Don W3FPR

> -----Original Message-----
>
> When I built my (second) K1-4 last year, it worked fine in all
> respects, with output as expected. Couple months ago it turned
> into a K1-2. I got full power on 40 and 30 M but less than 1/2 W
> on 20 and 15 M. I took it apart and measured RF voltages on 20
> M. They were low at all points. I put it together without the
> tuner. It worked on all bands. A K1-4 again. I put tuner back
> and it still worked fine. I used it without problem until today.
> Bact to a K1-2 again. 20 and 15 M had less than 1/2 W. I took
> it apart and removed tuner again. Big change. Now I get too
> much power on all bands. 10 W on 40 M to 5 W on 15 M. I have
> never got more than 7 W on any band. I turned power down to 0.1
> W and still got high power on all bands. Also, the RF indicator
> on the K1 dosn't work. No bars light. Looks like I have a power
> control problem. Like no control. Any ideas? Supply voltage was 12.7 V.
> 73
> Rick
> K7MW
>

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Rick Dettinger-2

Re: Need help for sick K1

Thanks Don - I will do that. I am using the K1 without the tuner now, to
see what happens for a while. Another list member reminded me to change the
power meter jumper on the RF board. That took care of the power request and
display problems. Now to find the 20 and 15 M power out problem. I am
having trouble with the problem being in the tuner as I don't see any band
specific circuitry there. I had assumed the problem would have to be in the
high freq. half of the filter board. Being intermittent greatly increases
the difficulty in troubleshooting. Banging the K1 on the table has no
effect. Neither does course language.
73
Rick
K7MW
----- Original Message -----
From: "Don Wilhelm" <[hidden email]>
To: "Rick Dettinger" <[hidden email]>; <[hidden email]>
Sent: Sunday, April 23, 2006 5:21 PM
Subject: RE: [Elecraft] Need help for sick K1

> Rick,
>
> Look carefully, then look again and again for the bad solder connection in
> the K1 tuner.
>
> 73,
> Don W3FPR
>
> > -----Original Message-----
> >
> > When I built my (second) K1-4 last year, it worked fine in all
> > respects, with output as expected. Couple months ago it turned
> > into a K1-2. I got full power on 40 and 30 M but less than 1/2 W
> > on 20 and 15 M. I took it apart and measured RF voltages on 20
> > M. They were low at all points. I put it together without the
> > tuner. It worked on all bands. A K1-4 again. I put tuner back
> > and it still worked fine. I used it without problem until today.
> > Bact to a K1-2 again. 20 and 15 M had less than 1/2 W. I took
> > it apart and removed tuner again. Big change. Now I get too
> > much power on all bands. 10 W on 40 M to 5 W on 15 M. I have
> > never got more than 7 W on any band. I turned power down to 0.1
> > W and still got high power on all bands. Also, the RF indicator
> > on the K1 dosn't work. No bars light. Looks like I have a power
> > control problem. Like no control. Any ideas? Supply voltage was 12.7

V.
> > 73
> > Rick
> > K7MW
> >

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Don Wilhelm-3

RE: Need help for sick K1

Rick,

If the problem persists with the tuner in bypass state, the most likely area
to look is the wattmeter section of the tuner.

73,
Don W3FPR

> -----Original Message-----
>
> Thanks Don - I will do that. I am using the K1 without the tuner now, to
> see what happens for a while. Another list member reminded me to
> change the
> power meter jumper on the RF board. That took care of the power
> request and
> display problems. Now to find the 20 and 15 M power out problem. I am
> having trouble with the problem being in the tuner as I don't see any band
> specific circuitry there. I had assumed the problem would have
> to be in the
> high freq. half of the filter board. Being intermittent greatly increases
> the difficulty in troubleshooting. Banging the K1 on the table has no
> effect. Neither does course language.
> 73
> Rick
> K7MW
> ----- Original Message -----
> From: "Don Wilhelm" <[hidden email]>
> To: "Rick Dettinger" <[hidden email]>; <[hidden email]>
> Sent: Sunday, April 23, 2006 5:21 PM
> Subject: RE: [Elecraft] Need help for sick K1
>
>
> > Rick,
> >
> > Look carefully, then look again and again for the bad solder
> connection in
> > the K1 tuner.
> >
> > 73,
> > Don W3FPR
> >
> > > -----Original Message-----
> > >
> > > When I built my (second) K1-4 last year, it worked fine in all
> > > respects, with output as expected. Couple months ago it turned
> > > into a K1-2. I got full power on 40 and 30 M but less than 1/2 W
> > > on 20 and 15 M. I took it apart and measured RF voltages on 20
> > > M. They were low at all points. I put it together without the
> > > tuner. It worked on all bands. A K1-4 again. I put tuner back
> > > and it still worked fine. I used it without problem until today.
> > > Bact to a K1-2 again. 20 and 15 M had less than 1/2 W. I took
> > > it apart and removed tuner again. Big change. Now I get too
> > > much power on all bands. 10 W on 40 M to 5 W on 15 M. I have
> > > never got more than 7 W on any band. I turned power down to 0.1
> > > W and still got high power on all bands. Also, the RF indicator
> > > on the K1 dosn't work. No bars light. Looks like I have a power
> > > control problem. Like no control. Any ideas? Supply
> voltage was 12.7
> V.
> > > 73
> > > Rick
> > > K7MW
>

Rick Dettinger-2

Re: Need help for sick K1

Thanks Don - It does. I was using a dummy load and I should have mentioned
that I bypassed the tuner in the menu.
73
Rick
K7MW
------------------------
> Rick,
>
> If the problem persists with the tuner in bypass state, the most likely
area
> to look is the wattmeter section of the tuner.
>
> 73,
> Don W3FPR

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Douglas Todd

K2 Calibration using WWV

Good Evening,

I've just completed soldering the K2, and have WWV at 10Mhz tuned in. The
display shows it at +380 hz high. According to the manual, on page 60, the
K2 can be calibrated using WWV. Unfortunately I do not see the procedure to
do this. Is it published ? Could I have a link to a a procedure on how to do
this ?

Thank you in advance.

Douglas Todd

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Ron D'Eau Claire-2

RE: K2 Calibration using WWV

I've just completed soldering the K2, and have WWV at 10Mhz tuned in. The
display shows it at +380 hz high. According to the manual, on page 60, the
K2 can be calibrated using WWV. Unfortunately I do not see the procedure to
do this. Is it published ? Could I have a link to a a procedure on how to do

this ?

Thank you in advance.

Douglas Todd

------------------------------

Yeah, it's sort of buried on the Elecraft Web site under "Notes on Adjusting
C22 to Put the K2 on Frequency".

I'll send you a PDF directly.

For any lurkers who are also curious, it's there on the web site somewhere
under "Construction Notes" I think.

The title says it all. It's not about "calibrating the dial", it's all about
putting the K2's internal clock on exactly 4.000 MHz (or as close as you
can) so that when you run CAL PLL and CAL FIL, the K2 in CAL FCTR mode
accurately measures the frequencies produced. Then the firmware can put the
rig more accurately on frequency when you set the dial.

The absolute accuracy is limited by a couple of factors. One is the
stability of the oscillators themselves. While crystal oscillators are used
for the BFO and the PLL reference oscillator, they still can drift slightly.
Indeed, both of those oscillators are designed to be adjusted in frequency.
The BFO frequency is adjusted to properly position the filter bandpass
during CAL FIL. The PLL Reference oscillator (the "local oscillator") is
actually tuned over a range to provide the nice smooth K2 tuning within each
'step' in the PLL. Whenever you make an oscillator that is supposed to
change frequency, it will naturally try to change frequency on its own. That
is, it will drift, at least somewhat. Finally the logic circuits limit how
accurate the dial calibration can be even if there was no drift at all in
the oscillators. When you tune in a frequency, the firmware generates a
certain digital value for the BFO and PLL Reference oscillators that puts
them on the right frequencies (it's these values that are measured and
stored during CAL PLL and CAL FIL). Logic circuits change these digital
values into analog voltages that actually control the oscillator
frequencies. There's a finite limit to the accuracy with which the
conversion from the digital value to the analog voltage can take place. The
residual error in the logic affects the dial accuracy.

Because of all those factors, the dial accuracy will typically be somewhere
within 30 Hz or so of the actual frequency, but may be off as much as 30 Hz
or so.

Ron AC7AC

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Tom Hammond-3

Re: K2 Calibration using WWV

In reply to this post by Douglas Todd

Douglas:

>I've just completed soldering the K2, and have WWV at 10Mhz tuned
>in. The display shows it at +380 hz high. According to the manual,
>on page 60, the K2 can be calibrated using WWV. Unfortunately I do
>not see the procedure to do this. Is it published ? Could I have a
>link to a a procedure on how to do this

Here's an 'improved' method for zeroing the MCU master Oscillator.
Comes directly from the designer, N6KR:

Subject: 4 MHz oscillator cal method for the K2
Date: Wed, 20 Aug 2003 13:57:44 -0700
From: Wayne Burdick <[hidden email]>

All of the methods that have been described for
calibrating the K2's 4-MHz oscillator will work. But I use
a different technique that allows C22 on the Control board
to be set to precisely to the correct position, with no
guesswork.

This method relies on the following simple observation. If
you tune in an on-air signal at a known frequency, the
difference between the *measured* VCO and BFO (using CAL
FCTR) *must* equal that frequency, or C22 is not set
correctly. (Actually, this holds for 160-17 m; on 15-10 m
it's the sum, or VCO + BFO, that must equal the signal's
frequency. But it's easier to do the adjustment of C22 on
17 m or lower because you don't have to do any math at
all, as I'll explain below.)

Here's the procedure. It requires revision 2.XX firmware,
and assumes you have already done Alignment and Test, Part
II, at some point. The K2 should also be allowed to come
up to room temperature.

1. Tune in a signal at a known frequency. Use one that's
at an *exact* kHz boundary, so you can easily see when
the VCO and BFO readings match in step 2. (I use WWV at
10, 15, or 20 MHz.) Use USB or LSB mode rather than CW,
so that there will be no CW receive offset. In the case
of a K2 I was calibrating, the VFO read 10000.17 when
the signal was tuned in perfectly. If it had read
10000.00, no further improvement would have been
possible.

TIP: Zero-beat the carrier precisely, or listen to a
voice signal and adjust the VFO for the best quality.
The more accurately you tune in the signal, the more
accurately you'll be able to set C22, below.

2. Run CAL FCTR. Now alternately move the K2's internal
counter probe between TP1 (VCO) and TP2 (BFO),
adjusting C22 in small increments until the kHz and Hz
digits at the two test points match as closely as
possible. In my case, the two readings matched at
14913.60 and 4913.60. The difference is exactly
10000.00--the frequency of the on-air signal.

3. Put the counter probe on TP1 (VCO), switch to 40
meters, and run CAL PLL.

4. Put the probe on TP2 (BFO) and run CAL FIL. For each
operating mode, vary each filter (or BFO) setting up 1
count, then back down, to force the K2 to take a new
BFO measurement for each and store it in EEPROM.

The VFO dial should now be very well calibrated.

If we get a lot of positive feedback on this method, we'll
post it as an application note.

73,
Wayne
N6KR

You might want to hold up on dong anything at this juncture, until
you get everything assembled and tested. It won't HURT to perform it
now, but you might need to re-do it at the end as wel..

73,

Tom Hammond N0SS

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Don Wilhelm-3

RE: K2 Calibration using WWV

Thanks for posting this Tom. It has since been written as an application
note and is available on the Elecraft website in .pdf format - see
http://www.elecraft.com/Apps/K2%20C22%20Adjust%20App%20Note.pdf

I have only one thing to add and that is for us 'tone-deaf' folks tuning WWV
precisely may be difficult (and the procedure does depend on precisely
tuning WWV or other reference frequency) - I can get close by listening to
the voice, but I may be as much as 100 Hz off.

I connect the computer running Spectrogram and listen to the transmitted
tones - WWV transmits 500 and 600 Hz tones on alternate minutes, and I look
for these tones displayed on the Spectrogram screen - I set markers at 500
Hz and 600 Hz - then adjust the tuning until the tones line up with the
markers. It is easy to get within 10 Hz this way.

Another way is to use the SPOT to provide a reference tone - set the
sidetone pitch at 500 or 600 Hz and tune the WWV tone for zero-beat with the
SPOT tone. Of course, if you zero with the 500 Hz tone thinking it was the
600 Hz one, you will be 100 Hz off, but listening for a short while will
quickly indicate this error.

There are other accurate ways of setting the 4 MHz master oscillator, and
some are contained in the K2 Dial Calibration article on my website
http://w3fpr.qrprqdio.com

73,
Don W3FPR

> -----Original Message-----
>
> Here's an 'improved' method for zeroing the MCU master Oscillator.
> Comes directly from the designer, N6KR:
>
> Subject: 4 MHz oscillator cal method for the K2
> Date: Wed, 20 Aug 2003 13:57:44 -0700
> From: Wayne Burdick <[hidden email]>
>
> All of the methods that have been described for
> calibrating the K2's 4-MHz oscillator will work. But I use
> a different technique that allows C22 on the Control board
> to be set to precisely to the correct position, with no
> guesswork.
>
> This method relies on the following simple observation. If
> you tune in an on-air signal at a known frequency, the
> difference between the *measured* VCO and BFO (using CAL
> FCTR) *must* equal that frequency, or C22 is not set
> correctly. (Actually, this holds for 160-17 m; on 15-10 m
> it's the sum, or VCO + BFO, that must equal the signal's
> frequency. But it's easier to do the adjustment of C22 on
> 17 m or lower because you don't have to do any math at
> all, as I'll explain below.)
>
> Here's the procedure. It requires revision 2.XX firmware,
> and assumes you have already done Alignment and Test, Part
> II, at some point. The K2 should also be allowed to come
> up to room temperature.
>
> 1. Tune in a signal at a known frequency. Use one that's
> at an *exact* kHz boundary, so you can easily see when
> the VCO and BFO readings match in step 2. (I use WWV at
> 10, 15, or 20 MHz.) Use USB or LSB mode rather than CW,
> so that there will be no CW receive offset. In the case
> of a K2 I was calibrating, the VFO read 10000.17 when
> the signal was tuned in perfectly. If it had read
> 10000.00, no further improvement would have been
> possible.
>
> TIP: Zero-beat the carrier precisely, or listen to a
> voice signal and adjust the VFO for the best quality.
> The more accurately you tune in the signal, the more
> accurately you'll be able to set C22, below.
>
> 2. Run CAL FCTR. Now alternately move the K2's internal
> counter probe between TP1 (VCO) and TP2 (BFO),
> adjusting C22 in small increments until the kHz and Hz
> digits at the two test points match as closely as
> possible. In my case, the two readings matched at
> 14913.60 and 4913.60. The difference is exactly
> 10000.00--the frequency of the on-air signal.
>
> 3. Put the counter probe on TP1 (VCO), switch to 40
> meters, and run CAL PLL.
>
> 4. Put the probe on TP2 (BFO) and run CAL FIL. For each
> operating mode, vary each filter (or BFO) setting up 1
> count, then back down, to force the K2 to take a new
> BFO measurement for each and store it in EEPROM.
>
> The VFO dial should now be very well calibrated.
>
> If we get a lot of positive feedback on this method, we'll
> post it as an application note.
>
> 73,
> Wayne
> N6KR
>
> You might want to hold up on dong anything at this juncture, until
> you get everything assembled and tested. It won't HURT to perform it
> now, but you might need to re-do it at the end as wel..
>
> 73,
>
> Tom Hammond N0SS
>
>
> _______________________________________________
> Elecraft mailing list
> Post to: [hidden email]
> You must be a subscriber to post to the list.
> Subscriber Info (Addr. Change, sub, unsub etc.):
> http://mailman.qth.net/mailman/listinfo/elecraft
>
> Help: http://mailman.qth.net/subscribers.htm
> Elecraft web page: http://www.elecraft.com
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.385 / Virus Database: 268.4.5/322 - Release Date: 4/22/2006
>
>

Dan Barker

RE: K2 Calibration using WWV

Watch out for the 440 Hz tones near the top of the hour. You also need a
clock<g>. The schedule is published.
(http://tf.nist.gov/stations/iform.html). Some minutes have NO tones.
Patience is required.

Dan / WG4S / K2 #2456
<snip>
I connect the computer running Spectrogram and listen to the transmitted
tones - WWV transmits 500 and 600 Hz tones on alternate minutes, and I look
for these tones displayed on the Spectrogram screen - I set markers at 500
Hz and 600 Hz - then adjust the tuning until the tones line up with the
markers. It is easy to get within 10 Hz this way.
</snip>

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Don Wilhelm-3

RE: K2 Calibration using WWV

I do this a LOT - and I spend the 'patience' time cleaning up the
workbench - but even that is not enough time, the workbench is still 'messy'
<G>

73,
Don W3FPR

> -----Original Message-----
> (http://tf.nist.gov/stations/iform.html). Some minutes have NO tones.
> Patience is required.
>

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Ron D'Eau Claire-2

RE: K2 Calibration using WWV

In reply to this post by Dan Barker

<snip>
I connect the computer running Spectrogram and listen to the transmitted
tones - WWV transmits 500 and 600 Hz tones on alternate minutes, and I look
for these tones displayed on the Spectrogram screen - I set markers at 500
Hz and 600 Hz - then adjust the tuning until the tones line up with the
markers. It is easy to get within 10 Hz this way. </snip>

--------------------------------------------------

Of course that's limited by the accuracy of the clock in your computer sound
card, which in some computers isn't very accurate!

Ron AC7AC

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Matt Osborn

Re: K2 Calibration using WWV

In reply to this post by Don Wilhelm-3

Don (and others)

WWV provides a BCD subcarrier at 100Hz that is always present. Using
HRD's PSK31 (it's free; thanks Simon!) it is a simple matter to adjust
the tuning to the proper frequency.

On Mon, 24 Apr 2006 10:01:19 -0400, "Don Wilhelm"
<[hidden email]> wrote:

>Thanks for posting this Tom. It has since been written as an application
>note and is available on the Elecraft website in .pdf format - see
>http://www.elecraft.com/Apps/K2%20C22%20Adjust%20App%20Note.pdf
>
>I have only one thing to add and that is for us 'tone-deaf' folks tuning WWV
>precisely may be difficult (and the procedure does depend on precisely
>tuning WWV or other reference frequency) - I can get close by listening to
>the voice, but I may be as much as 100 Hz off.
>
>I connect the computer running Spectrogram and listen to the transmitted
>tones - WWV transmits 500 and 600 Hz tones on alternate minutes, and I look
>for these tones displayed on the Spectrogram screen - I set markers at 500
>Hz and 600 Hz - then adjust the tuning until the tones line up with the
>markers. It is easy to get within 10 Hz this way.
>
>Another way is to use the SPOT to provide a reference tone - set the
>sidetone pitch at 500 or 600 Hz and tune the WWV tone for zero-beat with the
>SPOT tone. Of course, if you zero with the 500 Hz tone thinking it was the
>600 Hz one, you will be 100 Hz off, but listening for a short while will
>quickly indicate this error.
>
>There are other accurate ways of setting the 4 MHz master oscillator, and
>some are contained in the K2 Dial Calibration article on my website
>http://w3fpr.qrprqdio.com

Darrell Bellerive

Re: K2 Calibration using WWV

On April 24, 2006 01:56 pm, Matt Osborn wrote:
> WWV provides a BCD subcarrier at 100Hz that is always present. Using
> HRD's PSK31 (it's free; thanks Simon!) it is a simple matter to adjust
> the tuning to the proper frequency.

Please expand on this. I have never used a PSK application, but how would it
detect the subcarrier is exactly at 100 Hz? I suspect the software is using
the soundcard clock as a reference and that wouldn't necessarily be very
accurate.

Seems that the K2 can be set very exactly using Wayne's method
(http://www.elecraft.com/Apps/K2%20C22%20Adjust%20App%20Note.pdf) but it
requires properly zero beating WWV. Using an audio spectrum analyzer such as
Spectrogram to detect WWV's broadcast tones is dependant on the accuracy of
the spectrum analyzer.

The spot tone generator in the K2 seems to derive it's clock and therefore
it's frequency from the MCU and the 4.0 MHz oscillator set by C22, so the
spot tone frequency will be dependant on the setting of C22, which is what we
are trying to set in the first place. So using this tone as a reference does
not look like it will work.

Having a calibrated audio spectrum analyzer may be one solution. Perhaps a
musical tuning fork via a microphone could be used to calibrate the audio
spectrum analyzer. How accurate are tuning forks?

Wayne suggests just listening for zero beat, one should be able to get within
about 15 Hz.

The trick seems to be how to tune the K2 to exactly 10,000.000 kHz and at the
next step up of the vfo be at 10,000.010 kHz. Seems a lot like an ARRL
frequency measuring contest. :-)

What tricks have others on the list found to tune WWV to within 1 Hz or less
without expensive professionally calibrated test equipment?

73,
Darrell VA7TO K2#5093
--
Darrell Bellerive
Amateur Radio Stations VA7TO and VE7CLA
Grand Forks, British Columbia, Canada
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Nick Waterman

Re: K2 Calibration using WWV

Darrell Bellerive wrote:
> Please expand on this. I have never used a PSK application, but how would it
> detect the subcarrier is exactly at 100 Hz? I suspect the software is using
> the soundcard clock as a reference and that wouldn't necessarily be very
> accurate.

... but it'll be within a few percent, surely, and that means you'll be
within a few hertz of 100Hz? The K2 tuning steps are likely to be a lot
bigger than that?

> Seems that the K2 can be set very exactly using Wayne's method
> (http://www.elecraft.com/Apps/K2%20C22%20Adjust%20App%20Note.pdf) but it
> requires properly zero beating WWV.

It'll still be only within about 30Hz, won't it?

> Wayne suggests just listening for zero beat, one should be able to get within
> about 15 Hz.

... which I'd have thought is well within the soundcard clock accuracy too.

> What tricks have others on the list found to tune WWV to within 1 Hz or less
> without expensive professionally calibrated test equipment?

I think you'll find day-to-day temperature variation, and
capacitative/inductive effects of nearby objects probably cause more
variation than that :-)

--
"Nosey" Nick Waterman, G7RZQ, K2 #5209.
use Std::Disclaimer; [hidden email]
Confucius say: Man who fight with wife all day, get no piece at night
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Don Wilhelm-3

RE: K2 Calibration using WWV

In reply to this post by Darrell Bellerive

Darrell,

Since the limit of tuning precision for the K2 is 10 Hz, there is no benefit
to be derived by attempting to tune WWV or other standard frequency station
to less than 10 Hz - and that is well within the error range presented by a
computer soundcard. In other words, even if the error in the soundcard is 5
Hz, the tuning of the K2 will be as close as you can get.

Any error from a high frequency clock will be derived from a 'countdown' of
the computer clock or other crystal controlled clock. Even though a 1%
error on a 14 MHz clock is 140 kHz, when you divide it down to audio, the
same 1% error in the clock on a 100 Hz tone is only 1 Hz - and most computer
reference frequencies are much more accurate than the 1% I used as an
extreme example.

The same is true of the SPOT tone in the K2, it is derived from the 4 MHz
reference oscillator, and with just setting the C22 trimmer to midpoint the
frequency error will likely be less than 1 kHz - that is an error of one
part in 4000 which is only .025% error - when that same error is applied to
a 600 Hz tone, the error is only 0.15 Hz.

So it would seem to me that there is little actual need for calibration of
the computer soundcard spectrum analyzer. If the computer clock is so far
off that the soundcard is greatly in error, that fact should be obvious from
other 'funny happenings'.

73,
Don W3FPR

> -----Original Message-----

> Please expand on this. I have never used a PSK application, but
> how would it
> detect the subcarrier is exactly at 100 Hz? I suspect the
> software is using
> the soundcard clock as a reference and that wouldn't necessarily be very
> accurate.
>
> Seems that the K2 can be set very exactly using Wayne's method
> (http://www.elecraft.com/Apps/K2%20C22%20Adjust%20App%20Note.pdf) but it
> requires properly zero beating WWV. Using an audio spectrum
> analyzer such as
> Spectrogram to detect WWV's broadcast tones is dependant on the
> accuracy of
> the spectrum analyzer.
>
> The spot tone generator in the K2 seems to derive it's clock and
> therefore
> it's frequency from the MCU and the 4.0 MHz oscillator set by C22, so the
> spot tone frequency will be dependant on the setting of C22,
> which is what we
> are trying to set in the first place. So using this tone as a
> reference does
> not look like it will work.
>

Darrell Bellerive

Re: K2 Calibration using WWV

In reply to this post by Nick Waterman

Thanks Nick and Don for the explanation of the accuracy of the soundcards and
K2 sidetone.

On April 24, 2006 11:40 pm, Nick Waterman wrote:
> ... but it'll be within a few percent, surely, and that means you'll be
> within a few hertz of 100Hz? The K2 tuning steps are likely to be a lot
> bigger than that?

Theoretically if one could tune WWV to .001 Hz accuracy and have the manual
dexterity to set C22 as accurately, then at the frequency of 10,000,000.00 Hz
one would be within +/- .002 Hz. Now because the K2 does not tune in precise
10 hz steps, 10,000,010 Hz would only be as accurate as Wayne's algorithms
and PLL calibration allow. This is all assuming that there is absolutely no
change in frequency caused by temperature or any other means. Then again, WWV
is only accurate as received to +/- 1Hz (1 part in 10 million).

>
> I think you'll find day-to-day temperature variation, and
> capacitative/inductive effects of nearby objects probably cause more
> variation than that :-)

Yes, that is the reality of it.

Amazing how digital readout has changed the way we look at our frequency of
operation. Wonder what the typical accuracy of the digital readouts of the
commercial (Kenwood, Icom, Yaesu, Ten-Tec, etc.) is?

My first receiver was a Hallicrafters SX-140 and that was followed by a
Heathkit SB-101. One was dependant on the crystal calibrators for frequency
accuracy. The Hallicrafters used a 3.5 Mhz crystal to set the lower band
edges. There was no trimer to adjust the 3.5 MHz crystal. So at 3.5 MHz one
was as close to 3.5 MHz as the crystal oscillator tolerance. At the 4.0 MHz
end, it was dependant on the VFO and alignment. No spec was listed in the
manual for the frequency accuracy. Wonder what it was at 4.0 MHz, or at 54.0
MHz. 10 Hz, 100 Hz, no way. 10 Khz more likely.

Darrell

--
Darrell Bellerive
Amateur Radio Stations VA7TO and VE7CLA
Grand Forks, British Columbia, Canada
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Ron D'Eau Claire-2

RE: K2 Calibration using WWV

In addition to the inevitable drift in tunable oscillators, there is the
accuracy ("resolution" is a more common term) of the Digital-to-Analog
Converters (DACs) used in the K2 logic.

During CAL PLL and CAL FIL, the DACs must 'read' the analog tuning voltages
used to set the oscillators on frequency and convert it into a digital value
for storage. The DACs do this. They don't read the voltage to an infinite
number of places, so it gets 'rounded off'. The part that's rounded off is
lost and the value stored isn't exact.

When tuning around, the logic reads the tuning voltage data needed to put
the K2 on whatever frequency you've chosen from memory (they were stored
there during CAL PLL and CAL FIL) and runs the digital value through the
DACs again to produce the analog tuning voltages required by the varicaps in
the oscillators. Once again, the analog voltage produced by the DAC's is not
precise.

That's where most of the about 30 Hz ambiguity comes in. And it's why two
different K2's can vary in frequency by 30 Hz or so even if their
oscillators were absolutely accurate and drift free which, of course, they
are not.

Many folks new to the K2 don't realize that what they see on the frequency
display is NOT the actual frequency. It's the "requested" frequency chosen
by either punching in a frequency on the keypad or turning the dial. It does
not show any errors caused by oscillator drift or inaccuracy in the DACs.

The "plus" side of using tunable oscillators under logic control is that we
can easily set the filter bandpass positions using CAL FIL and the extremely
smooth tuning of the local oscillator with a very low level of phase noise.
The latter is the result of using not making the phase-locked loop step
constantly in very small frequency steps as you tune. The K2 PLL jumps in
fairly large steps. In between the large steps, the PLL Reference Oscillator
frequency is tuned to provide virtually continuous analog-style tuning.

Ron AC7AC

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

k6dgw

Re: K2 Calibration using WWV

Ron D'Eau Claire wrote:
> Many folks new to the K2 don't realize that what they see
> on the frequency display is NOT the actual frequency. It's
> the "requested" frequency chosen by either punching in a
> frequency on the keypad or turning the dial.

So ... running my K2 is like flying an F-16 or F-117A. I don't actually
"control" the aircraft or radio, I "request" the computer to control it
for me and hope it is listening?

I guess the DACs operate both ways (digital-->analog when controlling,
and analog-->digital) when calibrating. I calibrated my SX-28 from the
front panel -- first set the BANDSET knob and lock it down, then tune
about with the BANDSPREAD knob. I messed with the dial cord pullies and
WWV to initially get the BANDSET dial to line up at 10 Mcs. So far as I
know, the SX-28 did not have any DACs in it, and its read-outs were
decidedly non-digital.

I sold mine to raise a little money for college in mid-20th century. I
acquired another a few years ago from the family of W6ANX, and finally
donated it to the Western Historical Radio Museum. You can see a
picture of it at www.radioblvd.com/SX28Notes.html. Scroll down a
little, it's the one on the Hallicrafters speaker. The museum is a
really cool trip if you ever find yourself in Virginia City NV. A great
number of his items have been restored and work, like the E.F. Johnson
KW built into the desk and Woody's SX-28.

Fred K6DGW
Auburn CA CM98lw
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

Matt Osborn

Re: K2 Calibration using WWV

In reply to this post by Darrell Bellerive

There are three different signals on WWV 100, 400 & 500 Hz. I assumed
that if those signals had the proper relative frequencies that all I
had to do was check the absolute frequency. The Elecraft 2T-gen fit
that requirement with its two frequencies of 700 & 1900 Hz.

I prefer the PSK31 waterfall display to that of Spectrogram as the
trace readily shows any drift (there is none, but it's the warm fuzzy
I'm after).

On Mon, 24 Apr 2006 19:44:16 -0700, Darrell Bellerive <[hidden email]>
wrote:

>On April 24, 2006 01:56 pm, Matt Osborn wrote:
>> WWV provides a BCD subcarrier at 100Hz that is always present. Using
>> HRD's PSK31 (it's free; thanks Simon!) it is a simple matter to adjust
>> the tuning to the proper frequency.
>
>Please expand on this. I have never used a PSK application, but how would it
>detect the subcarrier is exactly at 100 Hz? I suspect the software is using
>the soundcard clock as a reference and that wouldn't necessarily be very
>accurate.
>
>Seems that the K2 can be set very exactly using Wayne's method
>(http://www.elecraft.com/Apps/K2%20C22%20Adjust%20App%20Note.pdf) but it
>requires properly zero beating WWV. Using an audio spectrum analyzer such as
>Spectrogram to detect WWV's broadcast tones is dependant on the accuracy of
>the spectrum analyzer.
>
>The spot tone generator in the K2 seems to derive it's clock and therefore
>it's frequency from the MCU and the 4.0 MHz oscillator set by C22, so the
>spot tone frequency will be dependant on the setting of C22, which is what we
>are trying to set in the first place. So using this tone as a reference does
>not look like it will work.
>
>Having a calibrated audio spectrum analyzer may be one solution. Perhaps a
>musical tuning fork via a microphone could be used to calibrate the audio
>spectrum analyzer. How accurate are tuning forks?
>
>Wayne suggests just listening for zero beat, one should be able to get within
>about 15 Hz.
>
>The trick seems to be how to tune the K2 to exactly 10,000.000 kHz and at the
>next step up of the vfo be at 10,000.010 kHz. Seems a lot like an ARRL
>frequency measuring contest. :-)
>
>What tricks have others on the list found to tune WWV to within 1 Hz or less
>without expensive professionally calibrated test equipment?
>
>73,
>Darrell VA7TO K2#5093